Around Christmas, I mounted a lighted star at the top of a radio
antenna tower.  The star was lit by a string of 35 2.5-V lamps in series
with 120 Volts across the string.  Just for fun and also to tell whether
anything seemed wrong with the lights, I read the resistance of the circuit
consisting of the bulbs and out-door extension cord that fed them.  It
occurred to me that one could read the outside temperature if one had a
very accurate bridge and calibrated the whole system.  With the lamps having
been off for several hours, the resistance was around 70 ohms when the
temperature was around 60 DG F.  We had one morning when it dropped to
11 or 12 DG and I seemed to recall that the resistance dropped to 67 Ohms.

        I don't have any figures on the coefficient of Tungsten or the other
parameters of the bulbs, but I wondered if I was seeing true temperature
variations or just a dirty socket on one of the bulbs.  When we got another
60 DG day, the resistance did go back up to around 69 or 70 Ohms.

        This whole idea might make a neat PIC project in that one could
run a small current through the string when it was supposed to be off and
sound an alarm if a bulb opened up.  Probably not worth it for Christmas lights,
but it might be for a safety light.

Martin McCormick WB5AGZ  Stillwater, OK 36.7N97.4W
OSU Center for Computing and Information Services Data Communications Group